|
|
||||||||
Regular Article |
Received March 14, 2001; revised January 30, 2002; accepted January 30, 2002. From the Neuropsychiatry Service, Department of Psychiatry and Behavioral Sciences, Johns Hopkins University, School of Medicine, Baltimore, MD (IL,JCSB,CGL), the Center for Mental Health and Department of Psychiatry, University of Athens, Athens, Greece (AV), and the Department of Mental Hygiene, Johns Hopkins University, School of Public Health, Baltimore, MD (JCSB,CGL). Address correspondence to Dr. Leroi, Osler 320, The Johns Hopkins Hospital, 600 North Wolfe St., Baltimore, MD 21287.
| ABSTRACT |
|---|
|
|
|---|
Key Words: Dementias (general) Psychosis Epidemiological Studies
| INTRODUCTION |
|---|
|
|
|---|
Delusions are unshakable false beliefs that are out of context with the person's social and cultural background.4 Hallucinations are false sensory perceptions that are not simply distortions or misinterpretations.5 Several misinterpretation delusions have been described in patients with dementia, including the following: a failure to recognize one's own home ("this house is not my home" phenomenon); beliefs that strangers are living in the house ("phantom boarder syndrome"); misinterpretations that one's loved ones are imposters (Capgras phenomenon6); beliefs that characters on the television are real and present in the room,7 and failure to recognize oneself in the mirror (the "mirror sign"8). Depending on their qualities, such misinterpretations might be regarded as delusions, abnormal visual phenomena, or manifestations of agnosia.
In spite of the frequency and importance of psychotic phenomena in dementia, there are few epidemiologic data on the prevalence of individual symptoms or on their occurrence in different dementing illnesses. Estimates from clinical samples have varied widely. Typically, these samples are from specialty-clinic series that are subject to referral bias. In a review of 21 studies, all but one from clinical samples, the reported prevalence of delusions in AD ranged from 10% to 73%.9 Likewise, the prevalence of hallucinations varied from 21% to 49%. Only two studies have examined the prevalence of misinterpretations separately,10,11 reporting prevalences of 23% and 50%. In addition to referral bias, reasons for these discrepancies include the use of different definitions for dementia and for individual psychotic phenomena, and lack of precision in small samples.
In the first large population study of psychosis in AD, Burns et al.12,13 examined 178 patients from a defined area in the United Kingdom. Among different sorts of delusions, they found ideas of theft to be most common (9%), especially in men, followed by delusions of suspicion (5.6%). Visual hallucinations (13%) were more common than auditory hallucinations (10%). Cognitive decline was more rapid in hallucinating (but not delusional) patients. The most common misidentification syndromesnotions that people were in the home (17.4%) and misidentification of others (11.8%)were more frequent in men and in those with earlier onset.
To date, the largest studies comparing symptom frequencies in various types of dementia rely on clinical samples. Thus, among 124 patients from four geriatric psychiatry services in the United Kingdom,14 delusions were most prevalent in Lewy body dementia (LBD; 75%), whereas visual hallucinations were common in both vascular dementia (VaD; 60%) and LBD (83.3%). Both were more common in older patients. Delusions were associated with deafness and recent difficult life events, whereas visual hallucinations were associated with visual impairment. By contrast, we have previously reported15 that the frequency of delusions was similar in outpatients with AD and VaD (29% and 22%, respectively), as was the frequency of hallucinations (12% and 15%).
Working in the Cache County cohort,16 we conducted a population study of psychiatric symptoms in dementia.17 Approximately 60% of individuals with dementia suffered from one or more mental or behavioral disturbances. Apathy, depression, and aggression/agitation were the most frequent psychiatric disturbances, but delusions were reported for 23% of those with AD and 8% of those with VaD. Hallucinations were reported in 13% of individuals with either AD or VaD. We did not describe individual psychotic symptoms, however, nor their severity or associations with other phenomena.
The present article describes a follow-up inquiry with the following aims: 1) to estimate the prevalence of individual psychotic phenomena in AD and VaD; 2) to compare their occurrence in AD and VaD; and 3) to evaluate the associations in AD between these psychotic phenomena and several sociodemographic or clinical variables.
| METHODS |
|---|
|
|
|---|
Sampling, Screening, and Clinical Assessment
We first screened participants for possible cognitive difficulty by use of the Modified Mini-Mental State Exam (3MS19). Mini-Mental State Exam (MMSE) scores were extracted from the 3MS, which is a more extensive cognitive assessment tool. The Dementia Questionnaire (DQ20) was then administered to knowledgeable informants of those whose initial screening scores suggested a possible cognitive syndrome and to informants for all participants age 90 or older. Participants with DQ ratings of 4 (suspected dementia) or 5 (probable dementia) were then invited to undergo a comprehensive Clinical Assessment (CA) at their residence (including nursing homes) in the presence of an informant. We also assigned a CA for a weighted, age-stratified probability sample of the entire population, regardless of screening results.
The CAs were conducted by experienced geriatric research teams. They included a history (from participants and informants), mental status examination, and the NPI, as well as a standardized neurologic examination, brief physical exam, and hour-long neuropsychological battery. Data from these evaluations, along with relevant medical records and lab results, were discussed at a conference of experts who classified study participants as "demented" or not, and assigned them (where appropriate) into specific diagnostic categories. We used standard criteria to differentiate Possible or Probable AD20 and Possible or Probable VaD.21
Assessment of Psychiatric Symptoms and Psychosis by Use of the Neuropsychiatric Inventory (NPI)
In all, 1,002 study participants received CAs, including NPI ratings. Some 201 of these individuals eventually received a diagnosis of AD, and 59, of VaD. We used the 10-item version of the NPI that divided mental/behavioral phenomena into 10 domains on the basis of observations within the past month. The NPI domains of Delusions and Hallucinations included ratings on nine individual forms of delusions and nine types of hallucinations. These two domains (but not the individual items) were rated on symptom frequency from 1 (occasionally) to 4 (very frequently, more than once a day). The domains were also rated on a 3-point severity scale (mild, moderate, severe) that denoted the degree to which they caused difficulty for the participants and their caregivers. We then multiplied these two ratings to obtain an overall domain rating of frequency and severity.
Analyses
We first identified participants with "any psychosis" (presence of delusions or hallucinations), and then noted those with only delusions or only hallucinations. Among participants with AD or VaD, we compared the proportions of those with scores of 1 or higher (presence of any disturbance(s) on domains of Delusion, Hallucination, or Any Psychosis). We assessed statistical significance of these comparisons by use of chi-square tests. We then compared mean combined scores (frequency x severity) for these same categories in AD versus VaD and assessed statistical significance with t-tests.
We compared the prevalence of individual types of delusions and hallucinations in the two dementia groups and also compared the frequency of two clusters of persecutory and misidentification delusions (chi-square or Fisher's exact tests). Finally, within the AD group, we similarly examined the association between several demographic and clinical risk factors and five psychosis categories: Any Psychosis, Delusions-alone, Hallucinations-alone, Persecutory Delusions, and Misidentification Delusions. Of note, missing data were relatively infrequent (less than 5%).
Taking into account the large number of variables being compared, a conservative p value (p<0.01) was chosen, with Bonferroni correction, for categories with no difference from "overall psychosis." For those in which a difference from the overall psychosis was found, a less restrictive p value (p <0.05) was used.
| RESULTS |
|---|
|
|
|---|
|
|
|
|
Table 5 shows demographic and clinical variables associated with psychotic symptoms in AD participants only (VaD numbers were too small to permit meaningful analysis). By chi-square analysis, women were seen to have more delusions, whereas those with a lower Clinical Dementia Rating scale (CDR) score (CDR Stage 1) had fewer delusions. Other variables, such as age at onset of AD, duration of illness, education level, presence of the apolipoprotein
4 allele, and MMSE score, did not appear to be associated with presence of psychosis or psychosis subtypes.
|
| DISCUSSION |
|---|
|
|
|---|
In AD, we found few cross-sectional associations for delusions or hallucinations. Delusions were associated with female gender, whereas an earlier stage of dementia (CDR Stage 1) was associated with fewer delusions. They were not associated with age at onset of AD, duration of illness, level of education, Apolipoprotein-E genotype, or with severity of dementia as gauged by the MMSE score.
Our finding of no difference in the rate of persecutory delusions between AD and VaD is consistent with the findings of Hirono et al.,23 Sultzer et al.,24 and Groves et al.15 These authors, in examining findings from clinical samples, reported no differences in the rates of psychosis between AD and VaD. This finding contrasts with what would be expected from a pathophysiological standpoint. As pointed out by Lyketsos et al.,17 the presence of persecutory delusions in AD may reflect the particular brain regions, namely the temporal lobes, affected by AD. Temporal lobe damage is more common in AD than in VaD and more likely to underlie psychotic symptoms, whereas in VaD, damage is more frequently seen in subcortical areas. Furthermore, in SPECT studies, a significant decrease in cerebral blood flow (CBF) can be seen in the inferior and superior temporal lobes25 of AD patients with delusions, as compared with those with no delusions.
Within the overall AD group (psychotic and nonpsychotic), the most prevalent symptom of the cluster of persecutory delusions was the belief that others were stealing from them (13%), followed by the belief of being in danger (7%; Table 3). These estimates are comparable to those by Burns et al.,12 who found delusions of theft to be the most common form of delusion subtype (9%) in their sample of 178 patients with AD, whereas jealous delusions (belief that one's spouse is having an affair) were seen in only 2% (Table 3) of all those with AD in our sample, and not at all in the VaD group. This low prevalence is in keeping with previous reports of delusional jealousy in dementia, which found rates to be between 2.3% and 5.6%.12,26,27 In contrast, Tsai et al.28 reported a relatively high rate of jealous delusions in AD (16.0%), perhaps because of selection bias. Delusional jealousy is particularly significant because the consequences of imagined infidelity could be harsh, for example, divorce.
The frequency of misidentification delusions was not significantly different between AD and VaD. This is consistent with Ballard et al.'s14 findings. Furthermore, that study found an increased prevalence of misidentification delusions with longer duration of illness. It is possible that cognitive impairment, particularly agnosia, is associated with misidentifications.
The estimates for individual hallucinations confirm previous work from clinical studies reporting that visual hallucinations are more frequent than auditory hallucinations in dementia.13 This finding, taken together with the profile of delusions in both AD and VaD, supports the argument presented by Jeste and Finkel29 that the psychosis seen in dementia, AD in particular, has phenomenology different from that in schizophrenia. The latter point is further supported by previous work from the CCSMA showing that neuropsychiatric disturbance in AD clusters into three groups, one of which is a psychotic group.30 Thus, the psychosis of AD appears to be a distinct syndrome, in which persecutory and misidentification delusions and visual hallucinations feature as the prominent symptoms.
We found few positive associations between clinical and demographic variables and the prevalence of psychosis in AD, other than a positive association between psychosis overall, delusions and persecutory delusions, and female gender. This is consistent with Rockwell et al.'s findings.31 However, in the UK population study, Burns et al.12 reported a higher prevalence of delusions in men. This discrepancy may be an artifact, although it is tempting to suggest that it might reflect cultural differences between American and British patients.
Among AD participants, we found no difference in the prevalence of delusions with increasing severity of cognitive impairment as assessed on the MMSE. However, a lower prevalence of delusions was seen in the earlier stages of dementia, as measured by the CDR. Delusions have been reported in all stages of dementia,2,31,32 with at least one study finding delusions to be more prevalent during the moderate stage.33 Cooper et al.34 described the relationship with noncognitive behavioral dysfunction as quadratic in nature, with fewer occurrences of psychosis in the early and later stages of the illness. Our results do not support this pattern, possibly because of our examination of psychosis as an entity separate from noncognitive behavioral disturbances as a whole.
There were no correlating clinical and demographic variables for hallucinations. This is consistent with previous work from our group reporting that delusions and hallucinations in AD have different correlates. However, in contrast to our current findings, Bassiony et al.35 found that more severe dementia is a principal risk factor for hallucinosis. Others have found associations between visual hallucinations in AD and older age and sensory deficits such as low visual acuity, as well as the presence of visual agnosia36 and occipital lobe atrophy.37 Our study did not address these associations further. On the basis of these findings, delusions and hallucinations in AD may have different pathophysiologic mechanisms. For example, it has been suggested that visual hallucinations may be due to cholinergic deficits in the parietal and temporal lobes.38
Finally, we found no association between the number of APOE
4 alleles in AD patients and an increased occurrence of psychotic symptoms. This is supported by previous findings of no association between AD patients carrying the APOE
4 allele and an elevated risk for psychosis.39,40 However, results regarding such an association have been inconclusive, and some reports have supported an association.4143 The last study found the positive association only for patients in the severe stage of AD, after controlling for the effects of age, gender, education, and level of cognitive impairment.43
Limitations to the current study lie in the complex sampling design. Such a weighted, age-stratified probability sample of an entire population may introduce inaccuracies in standard errors. This, however, is unlikely to have a large impact on these results. Another limitation of the study is that, although it is a population-based epidemiologic study, the population sampled may not be representative of others. That is, the population in Cache County is older and less ethnically diverse than elsewhere in the United States. The use of the NPI is limited in that it relies on informant reports for ascertaining behavioral disturbances, rather than on direct patient assessment. Also, although the NPI raters were trained in the standard methods required by the NPI, no interrater reliability was established.
In summary, we present estimates of the prevalence and severity of psychotic symptoms in AD and VaD. The data are from a large population-based survey, and therefore this study avoids the pitfall of selection bias that so often bedevils work in this area. Of particular interest is the finding that the profile of delusions and hallucinations seen is different from that seen in schizophrenia, further supporting the hypothesis that AD-associated psychosis is a distinct phenomenological syndrome. Continued close examination of these types of phenomena may shed further light on the pathogenesis of delusions and hallucinations and might further advance our ability to treat these troubling symptoms in patients with dementia.
| ACKNOWLEDGMENTS |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Wiederkehr, M. Simard, C. Fortin, and R. van Reekum Validity of the Clinical Diagnostic Criteria for Vascular Dementia: A Critical Review. Part II J Neuropsychiatry Clin Neurosci, May 1, 2008; 20(2): 162 - 177. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. T. Beier Pharmacotherapy for behavioral and psychological symptoms of dementia in the elderly Am. J. Health Syst. Pharm., January 15, 2007; 64(2_Supplement_1): S9 - S17. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Borroni, M. Grassi, C. Agosti, S. Archetti, C. Costanzi, C. Cornali, C. Caltagirone, L. Caimi, M. Di Luca, and A. Padovani Cumulative Effect of COMT and 5-HTTLPR Polymorphisms and Their Interaction With Disease Severity and Comorbidities on the Risk of Psychosis in Alzheimer Disease Am J Geriatr Psychiatry, April 1, 2006; 14(4): 343 - 351. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Ropacki and D. V. Jeste Epidemiology of and Risk Factors for Psychosis of Alzheimer's Disease: A Review of 55 Studies Published From 1990 to 2003 Am J Psychiatry, November 1, 2005; 162(11): 2022 - 2030. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ALL ISSUES | SEARCH | TABLE OF CONTENTS |